Independent · Specification-led · No sponsored recommendations
Independent AI hardware reviews, benchmarks and buying guides
Find the right hardware for local AI, LLMs, development and your homelab — chosen by what it can actually run, not by what its spec sheet claims.
Published scoring methodology Basis of every assessment stated on the page No paid placements or paid scores
Start with your hardware
Eight pillars, each with the guides, reviews and comparisons that go with them.
- AI Workstations Complete desktop and tower systems for local AI, model fine-tuning and heavy development workloads — chosen by what…
- GPUs Graphics cards evaluated for AI work, where VRAM capacity and memory bandwidth usually matter far more than raw…
- Mini PCs Small-form-factor systems for homelabs, always-on inference nodes and quiet desk setups, including the new generation of unified-memory machines.
- Local AI Hardware requirements, memory maths and capability guides for running language and image models on your own machine.
- Homelab Servers, racks, power and the supporting infrastructure that keeps a self-hosted lab running reliably.
- Storage SSDs, NAS units and dataset storage sized for model weights, checkpoints and training data.
- Networking Switches, routers and adapters for moving large models and datasets around a lab without waiting on gigabit.
- Developer Gear Monitors, keyboards, docks and desk hardware chosen for people who spend the working day in an editor and…
Buying guides
Recommendations organised by what you are trying to run, with the reasoning shown.
- Buying Guide Best GPUs Under $1,000 for AI The best card at this budget is usually not a current one. New tops out at 16 GB; used reaches 24 —…
- Buying Guide Best NAS for AI Developers A NAS for model weights is optimised differently from one for media — throughput, cache and memory matter, and the transcoding engine…
- Buying Guide Best NAS Drives for AI Datasets The enclosure is the visible decision; the drives determine whether the array survives its first failure. CMR, workload ratings, and the batch-diversity…
- Buying Guide Best Mini PCs for Developers Sustained multi-core performance is what makes a development machine feel fast — and configured TDP, which no listing publishes, is why two…
- Buying Guide Best AI Workstations Under $2,000 What this budget actually reaches — 16 GB comfortably, 24 GB used — and the one thing it does not buy at…
- Buying Guide Best KVM Switches for Developers and Homelabs Check your monitor first — many already have one. If not: two different products share this name, and bandwidth and EDID handling…
Local AI
The memory arithmetic and hardware requirements behind running models on your own machine.
- Explainer Stable Diffusion GPU Requirements What image generation actually needs, model by model — and why the answer changed when the 12-billion-parameter models arrived.
- Explainer LM Studio Hardware Requirements The requirements belong to the model, not the app — but LM Studio hands you the GPU offload slider Ollama hides, which…
- Comparison 64GB vs 128GB RAM for Local AI: Which Do You Need? The answer depends entirely on whether your GPU has its own memory. With a discrete card, 64 GB is ample. With unified…
Popular comparisons
Head-to-head, with a clear answer about which one you should actually buy.
- Comparison NAS vs DAS for AI Storage DAS is faster and simpler; NAS is shareable and tells you when a drive dies. Which matters depends on whether one machine…
- Comparison Mini PC vs Desktop for Local AI A form-factor decision with a counter-intuitive consequence: the small machine runs larger models than the large one, and the large one runs…
- Comparison Mac Studio vs AI PC for Local LLMs A calculable trade, not a platform preference. The crossover is 32 GB — below it the PC wins on speed, above it…
- Comparison 2.5GbE vs 10GbE: Which Do You Actually Need? You probably already own 2.5-gigabit ports and the cabling for them. Whether 10 gigabit is worth the upgrade comes down to one…
- Comparison 64GB vs 128GB RAM for Local AI: Which Do You Need? The answer depends entirely on whether your GPU has its own memory. With a discrete card, 64 GB is ample. With unified…
- Comparison NVIDIA vs AMD for Local AI: What the Gap Actually Costs AMD's hardware is competitive and its software is not — but how much that matters depends entirely on whether you are generating…
AI workstations
Complete systems, assessed by the models they can hold in memory.
- Build The Machine You Already Own: A Reference Architecture Every other build here starts from a blank sheet. This one starts from an inventory — find the one component actually stopping…
- Build The Reproducible AI Machine: A Reference Architecture Same model, same prompt, same seed, same GPU — and still a different answer. Why that happens, when it matters, and the…
- Build The Dual-Purpose AI and Gaming PC: A Reference Architecture This site tells you to ignore gaming rankings. If you also game, you cannot — the two workloads want almost disjoint parts…
GPUs for AI
Graphics cards evaluated on VRAM and memory bandwidth, not frame rates.
- Review AMD Radeon RX 9070 XT Review: A Better Card That Runs Models Worse Newer, more efficient and architecturally improved — with 50% less memory and 49% less bandwidth than the card it follows. For local…
- Review AMD Radeon RX 7900 XTX Review: The Hardware Is Not the Problem The only new card near this price with 24 GB, and it beats both 16 GB NVIDIA options on the specifications that…
- Review NVIDIA RTX PRO 6000 Blackwell Review: Not Faster, Just Able The same memory bandwidth as an RTX 5090 with three times the capacity — so you are not buying speed. You are…
Mini PCs
Small machines for homelabs, inference nodes and quiet desks.
- Buying Guide Best Mini PCs for Developers Sustained multi-core performance is what makes a development machine feel fast — and configured TDP, which no listing publishes, is why two…
- Comparison Mini PC vs Desktop for Local AI A form-factor decision with a counter-intuitive consequence: the small machine runs larger models than the large one, and the large one runs…
- Buying Guide Best Mini PCs for Local LLMs Unified-memory mini PCs run 70B models no consumer graphics card can load. Which ones are worth buying, what they cost you in…
Latest reviews
Scored against our published methodology, with the basis of assessment stated.
- Review AMD Radeon RX 9070 XT Review: A Better Card That Runs Models Worse Newer, more efficient and architecturally improved — with 50% less memory and 49% less bandwidth than the card it follows. For local…
- Review AMD Radeon RX 7900 XTX Review: The Hardware Is Not the Problem The only new card near this price with 24 GB, and it beats both 16 GB NVIDIA options on the specifications that…
- Review NVIDIA RTX PRO 6000 Blackwell Review: Not Faster, Just Able The same memory bandwidth as an RTX 5090 with three times the capacity — so you are not buying speed. You are…
Homelab and infrastructure
Storage, networking and the supporting hardware that keeps a lab running.
- Build The Multi-Tenant GPU Machine: A Reference Architecture The multi-user server assumes its users are colleagues. Remove that assumption and almost everything changes — consumer cards stop qualifying, and prefix…
- Build The Colocated AI Machine: A Reference Architecture A machine nobody can physically touch is specified around its recovery ladder. Out-of-band management, network-bound unlock, watchdogs and the runbook a stranger…
- Build The Multi-User Inference Server: A Reference Architecture Serving many people inverts the arithmetic: batching makes concurrency nearly free, and the KV cache — not the model — is what…
Developer gear
Monitors, keyboards and docks chosen for people who spend the day in an editor.
- Buying Guide Best KVM Switches for Developers and Homelabs Check your monitor first — many already have one. If not: two different products share this name, and bandwidth and EDID handling…
- Buying Guide Best Monitors for Developers Pixel density decides whether text is comfortable all day — and the right target depends on whether you are on macOS, Windows…